NEXIDIA'S LANGUAGE IDENTIFICATION

Nexidia’s Language Identification (LID) tool gives users an automated way to identify and categorize the spoken language used in an audio segment by providing the ability to rapidly distinguish between disparate languages and dialects.

Language Identification offers the means to automatically segregate audio/video data into identified languages with an extremely high degree of confidence. This enables end users to only receive the language(s) or custom language model that is most relevant to them, not only increasing productivity but ensuring the file is of a profile that warrants further evaluation and is being evaluated by the proper resource.

LID has the ability to be trained on any language, dialect or customized language model that may be of interest. A customized model is developed to identify a language model particular to an organizations needs, e.g. LID can be trained to classify recorded interviews into acceptable audio and unacceptable audio for the purpose of hiring contact center agents. Once trained, the tool can then be used to automatically identify those languages or models as identified by the organization. The language identification engine allows the user to supply an arbitrary segment of audio in any supported format so that the Language Identification Model (LIM) can then identify which language was most likely spoken in the audio segment. The result will provide the language or language model along with the filename of the given file. The ability to quickly segment an audio clip into the correct language, dialect or custom language model is used to quickly and accurately segment audio as a means of replacing the largely manual process currently being used by government and commercial organizations. As a result these organizations are benefiting from improved and quicker decision making; reduced overhead and increased productivity.